Easy experimenting and tracking with the deep learning stack deployed in containers.
In the previous post, we set up a multi-container application with Docker Compose, where one container is responsible for hosting a Jupyter Notebook environment, and another container holds a Tensorboard deployment. Containers are going to communicate via directories that are mounted as volumes from the host machine.
Today’s primary goal is to see how this setup works and how it enables easy experiment tracking. For this, we will start working towards creating a news recommender system that will be (obviously) backed by a neural network. As with any recommender system, it will propose top-10 most similar news, and a neural…
TL;DR
Containers are lightweight software packages that run in isolation on the host computing environment. Using containerized environments for the research is helpful because they are easy to use, reproducible, and keep your primary system clean. You can easily set up a fully functioning multi-container application using Docker Compose. All scripts required for this are available here.
Update 02/2021: the second part of the series is available here.
Why containers?
When doing intensive research, sometimes it becomes hard to maintain what is going on around you. Tons of files with data, dozens of libraries that you don’t use but not sure enough to…
